Search

Forward–backward doubly stochastic differential equations (FBDSDEs) serve as a probabilistic interpretation of stochastic partial differential equations (SPDEs) with diverse applications. Coupled FBDSDEs encounter numerous challenges in numerical approximation compared to forward–backward stochastic differential equations (FBSDEs) and decoupled FBDSDEs, including ensuring the measurability of the numerical solutions, accounting for the mutual influences between forward and backward processes, and considering the relationship with respect to SPDEs rather than PDEs. This paper introduces, for the first time, a numerical method for solving multi-dimensional coupled FBDSDEs. By integrating an optimal control-based approach with deep neural networks, it effectively addresses the coupling-related challenges between forward and backward equations. Computational examples of coupled FBDSDEs with explicit solutions demonstrate that the proposed deep learning-based numerical algorithm achieves commendable performance in terms of both accuracy and efficiency. [ABSTRACT FROM AUTHOR]

This paper employs a deep reinforcement learning algorithm named Twin Delayed Deep Deterministic algorithm into autonomous navigation in intelligent transportation systems. It trains a fully connected neural network model in a simulation environment, which outputs the expected linear and angular velocity of the vehicle based on real-time data measured by embedded sensors. Through continuous epochs of training, the model gradually navigates the vehicle to reach a provided destination by making rational motion decisions at each discrete time instant without knowing global environment information. Especially, to improve the model's generalization ability across various scenes, an input preprocessing function is proposed to eliminate the singularity and uniformity of raw input data. A large number of simulation tests are carried out, where the proportion that the vehicle moves from a start position to a destination without collision within a specified limited time exceeds 90%. The remaining failures are mainly due to the vehicle's inability to approach the destination immediately adjacent to obstacles for its safety. Furthermore, traditional mapless navigation algorithms suffer from locally optimal solutions in the face of U-shaped obstacles. This paper introduces a virtual obstacle mechanism designed to prevent the vehicle from entering the U-shaped region, effectively addressing the aforementioned issue. Finally, the model trained from the simulation environment can be directly loaded onto a physical vehicle without considering the different processor architectures. Large quantities of experiments show that the model improves the autonomous navigation capability of vehicles when global environment information cannot be obtained by the system, which optimizes the functions of the navigation module in intelligent transportation systems. • The TD3 navigation algorithm's custom parameters achieve superior performance. • A preprocessing function enhances the navigation model's generalization capability. • A virtual obstacle mechanism effectively addresses the issue of U-shaped obstacles. • The model trained in simulation can be directly deployed on physical vehicles. [ABSTRACT FROM AUTHOR]

This paper explores methods to reduce aircraft design loads through an optimal fuel usage strategy, which maximises the beneficial effects of wing-tank sloshing-induced damping. A reduced-order neural network models the nonlinear sloshing behaviour in various tank fill level scenarios. Integrated into an aeroelastic framework, this model allows for the assessment of incremental sloshing-induced damping in gust response of wings. The optimised fuel usage strategy enables control of fuel consumption from each individual tank to provide maximisation of load alleviation. • The paper presents the development of a reduced order model based on neural networks to describe vertical sloshing at different fill levels. • The model was validated by reproducing a single degree of freedom experiment at different tank scales. • The model was fed into an aeroelastic simulation framework using the Goland wing as a benchmark. • A fuel level control strategy was developed to minimise wing root loads through an optimisation process. [ABSTRACT FROM AUTHOR]

In wind speed prediction technologies, deep learning-based methods have achieved promising advantages. However, most existing methods focus on learning implicit knowledge in a data-driven manner but neglect some explicit knowledge from the physical theory of meteorological dynamics, failing to make stable and long-term predictions. In this paper, we explore introducing explicit physical knowledge into neural networks and propose Physical Equations Predictive Network (PEPNet) for multi-step wind speed predictions. In PEPNet, a new neural block called the Augmented Neural Barotropic Equations (ANBE) block is designed as its key component, which aims to capture the wind dynamics by combining barotropic primitive equations and deep neural networks. Specifically, the ANBE block adopts a two-branch structure to model wind dynamics, where one branch is physic-based and the other is data-driven-based. The physic-based branch constructs temporal partial derivatives of meteorological elements (including u-component wind, v-component wind, and geopotential height) in a new Neural Barotropic Equations Unit (NBEU). The NBEU is developed based on the barotropic primitive equations mode in numerical weather prediction (NWP). Besides, considering that the barotropic primitive mode is a crude assumption of atmospheric motion, another data-driven-based branch is developed in the ANBE block, which aims at capturing meteorological dynamics beyond barotropic primitive equations. Finally, the PEPNet follows a time-variant structure to enhance the model's capability to capture wind dynamics over time. To evaluate the predictive performance of PEPNet, we have conducted several experiments on two real-world datasets. Experimental results show that the proposed method outperforms the state-of-the-art techniques and achieve optimal performance. [ABSTRACT FROM AUTHOR]

The problem of vanishing and exploding gradients has been a long-standing obstacle that hinders the effective training of neural networks. Despite various tricks and techniques that have been employed to alleviate the problem in practice, there still lacks satisfactory theories or provable solutions. In this paper, we address the problem from the perspective of high-dimensional probability theory. We provide a rigorous result that shows, under mild conditions, how the vanishing/exploding gradients problem disappears with high probability if the neural networks have sufficient width. Our main idea is to constrain both forward and backward signal propagation in a nonlinear neural network through a new class of activation functions, namely Gaussian–Poincaré normalized functions, and orthogonal weight matrices. Experiments on both synthetic and real-world data validate our theory and confirm its effectiveness on very deep neural networks when applied in practice. • The vanishing/exploding gradient problem in training deep neural networks is addressed. • The problem is provably solved under mild conditions using high-dimensional probability theory. • Experiments show neural network of 200 layers can be trained without linearizing the networks. [ABSTRACT FROM AUTHOR]

Weight-tied models have attracted attention in the modern development of neural networks. The deep equilibrium model (DEQ) represents infinitely deep neural networks with weight-tying, and recent studies have shown the potential of this type of approach. DEQs are needed to iteratively solve root-finding problems in training and are built on the assumption that the underlying dynamics determined by the models converge to a fixed point. In this paper, we present the stable invariant model (SIM) , a new class of deep models that in principle approximates DEQs under stability and extends the dynamics to more general ones converging to an invariant set (not restricted in a fixed point). The key ingredient in deriving SIMs is a representation of the dynamics with the spectra of the Koopman and Perron–Frobenius operators. This perspective approximately reveals stable dynamics with DEQs and then derives two variants of SIMs. We also propose an implementation of SIMs that can be learned in the same way as feedforward models. We illustrate the empirical performance of SIMs with experiments and demonstrate that SIMs achieve comparative or superior performance against DEQs in several learning tasks. • A Koopman spectral analysis is performed for deep equilibrium models. • The analysis identifies stable dynamics determined by deep equilibrium models. • We develop a class of deep models to approximate the stable dynamics. • Models in the class can represent dynamics converging to an invariant set. [ABSTRACT FROM AUTHOR]

The application of neural network models to solve combinatorial optimization has recently drawn much attention and shown promising results in dealing with similar problems, like Travelling Salesman Problem. The neural network allows to learn solutions based on given problem instances, using reinforcement learning or supervised learning. In this paper, we present a novel end-to-end method to solve routing problems. In specific, we propose a gated cosine-based attention model (GCAM) to train policies, which accelerates the training process and the convergence of policy. Extensive experiments on different scale of routing problems show that the proposed method can achieve faster convergence of the training process than the state-of-the-art deep learning models while achieving solutions of the same quality. • We proposed a cosine-based neural network model based on the transformer architecture to improve node embeddings. • We use cosine decomposability to accelerate training speed of the proposed model. • We use a gating connection to stabilize training and get faster convergence, which leads to better overall performance. • We apply the proposed gated cosine-based attention model to solve several Routing Problems. [ABSTRACT FROM AUTHOR]

To tune the wave energy converter (WEC) controller parameters such as damping to reduce the line force during extreme wave conditions, future knowledge of the line force is required. To achieve this, the incoming wave and system state should be predicted for a few seconds in the future. It is rather an arduous task to predict the future knowledge of waves and the system's dynamic when dealing with breaking and steep waves, and the system is subject to various nonlinear forces. The classical model-based control strategies often rely on linear assumptions to estimate the WEC dynamics for the sake of simplicity. Unlike the model-based, the data-driven approaches are free from modeling errors and the algorithms are trained over the true and noisy data to predict non-linear system behaviors. Using data-driven approaches, we are able to model nonlinear dynamics. However, new questions emerge on the accuracy of the future wave and system state predictions, and how this uncertainty propagates into the final prediction of the line force. As incorrect damping may lead to excessive line force and detrimental damage to the system, these are the knowledge gaps that need to be addressed. The main purpose of this paper is to answer these questions through a survivability strategy for wave energy converters by providing a realistic perspective on the implementation of the neural network approaches by accounting for the errors in the input data. For this purpose, a series of neural networks is designed that first predicts the surface elevation for 0.36 s ahead, i.e. corresponding to 2 s in the full-scale WEC. This future knowledge of the wave elevation is then used to predict the system state (i.e. power take-off (PTO) translator position) for the same prediction horizon based on the PTO damping. This information is then fed to a convolutional neural network (CNN) that predicts the peak line force 0.36 s ahead. This paper also evaluates the predictive capabilities of neural network (NN) models, comparing them to classical autoregressive (AR) and Kalman filter methods. While the AR model exhibits slightly higher accuracy in fixed PTO system configurations, it falters in providing real-time predictions when system parameters undergo variations. In contrast, the NN prediction model excels by establishing a robust relationship between system configuration and output signals. Especially noteworthy is its ability to outperform classical methods with minimal training on a selective set of system configurations. Then, the sensitivity of the peak line force prediction to the uncertainties in the input data from NN models and the prediction horizon is analyzed. The neural network models are trained over the experimental data subjected to extreme sea states for a point absorber wave energy converter. The results suggest that the accuracy of the surface elevation prediction has an insignificant direct effect on the peak force prediction model. However, these uncertainties are reflected in the PTO translator position prediction, and the model is considerably sensitive to the accuracy of this prediction. This sensitivity nonetheless is less notable for higher PTO damping values. • Predicting wave elevation, PTO position, and mooring force via neural networks. • Finding optimum damping to reduce mooring force in extreme wave conditions. • Predicting the mooring force is sensitive to the PTO position prediction. [ABSTRACT FROM AUTHOR]

As a powerful data modeling tool, Heterogeneous Information Network (HIN) has been successfully used in auxiliary information exploitation to boost recommendation performance. For HIN based recommendation, it is challenging to extract and fuse useful features of user preferences and item attributes under different semantic paths in HINs. Existing methods leverage a pre-defined fusion function to integrate different semantics for recommendation, which cannot characterize the complex nonlinear interactions between users and items. In this paper, we present a general framework named MNRec, short for Meta-path fusion based Neural Recommendation, to extract and fuse user and item embeddings under different meta-paths for recommendation. Under the framework, we propose an instantiation of MNRec with Multi-Layer Perceptron (MLP) structure. It consists of two major steps, i.e., meta-path based heterogeneous network embedding and deep learning based rating prediction. Concretely, appropriate meta-paths are first designed according to domain knowledge. Then the embeddings of users and items are obtained through a meta-path and commuting matrix based heterogeneous network embedding method. Finally, in light of the powerful nonlinear modeling capabilities of deep neural networks, the learned embeddings under different meta-paths are integrated into a two-pathway MLP structure for rating prediction. Experimental results on three real-world datasets demonstrate the superiority and effectiveness of MNRec compared with state-of-the-art baselines in rating prediction. [ABSTRACT FROM AUTHOR]

As the first review in this field, this paper presents an in-depth mathematical view of Intelligent Flight Control Systems (IFCSs), particularly those based on artificial neural networks. The rapid evolution of IFCSs in the last two decades in both the methodological and technical aspects necessitates a comprehensive view of them to better demonstrate the current stage and the crucial remaining steps towards developing a truly intelligent flight management unit. To this end, in this paper, we will provide a detailed mathematical view of Neural Network (NN)-based flight control systems and the challenging problems that still remain. The paper will cover both the model-based and model-free IFCSs. The model-based methods consist of the basic feedback error learning scheme, the pseudocontrol strategy, and the neural backstepping method. Besides, different approaches to analyze the closed-loop stability in IFCSs, their requirements, and their limitations will be discussed in detail. Various supplementary features, which can be integrated with a basic IFCS such as the fault-tolerance capability, the consideration of system constraints, and the combination of NNs with other robust and adaptive elements like disturbance observers, would be covered, as well. On the other hand, concerning model-free flight controllers, both the indirect and direct adaptive control systems including indirect adaptive control using NN-based system identification, the approximate dynamic programming using NN, and the reinforcement learning-based adaptive optimal control will be carefully addressed. Finally, by demonstrating a well-organized view of the current stage in the development of IFCSs, the challenging issues, which are critical to be addressed in the future, are thoroughly identified. As a result, this paper can be considered as a comprehensive road map for all researchers interested in the design and development of intelligent control systems, particularly in the field of aerospace applications. [ABSTRACT FROM AUTHOR]

Deep Neural Networks are widely used for solving complex problems in several scientific areas, such as speech recognition, machine translation, image analysis. The strategies employed to investigate their theoretical properties mainly rely on Euclidean geometry, but in the last years new approaches based on Riemannian geometry have been developed. Motivated by some open problems, we study a particular sequence of maps between manifolds, with the last manifold of the sequence equipped with a Riemannian metric. We investigate the structures induced through pullbacks on the other manifolds of the sequence and on some related quotients. In particular, we show that the pullbacks of the final Riemannian metric to any manifolds of the sequence is a degenerate Riemannian metric inducing a structure of pseudometric space. We prove that the Kolmogorov quotient of this pseudometric space yields a smooth manifold, which is the base space of a particular vertical bundle. We investigate the theoretical properties of the maps of such sequence, eventually we focus on the case of maps between manifolds implementing neural networks of practical interest and we present some applications of the geometric framework we introduced in the first part of the paper. • We consider Neural Networks as sequences of smooth maps between manifolds. • We investigate the structures induced through pull-backs on the manifolds. • The pull-backs of the final Riemannian metric to any manifolds is a degenerate metric. • The theoretical study leads to construct equivalence classes in the input manifold. [ABSTRACT FROM AUTHOR]

Since input constraints and external disturbances are unavoidable in tracking control problems, how to obtain a controller in this case to save communication and data resources at the same time is very challenging. Aiming at these challenges, this paper develops a novel neural network (NN)-based event-triggered integral reinforcement learning (IRL) algorithm for constrained H ∞ tracking control problems. First, the constrained H ∞ tracking control problem is transformed into a regulation problem. Second, an event-triggered optimal controller is designed to reduce network transmission burden and improve resource utilization, where a novel threshold is proposed and its non-negativity can be guaranteed. Third, for implementation purpose, a novel NN-based event-triggered IRL algorithm is developed. In order to improve data utilization, the experience replay technique with an easy-to-verify condition is employed in the learning process. Theoretical analysis proves that the tracking error and weight estimation error are uniformly ultimately bounded. Finally, simulation verification shows the effectiveness of the present method. [ABSTRACT FROM AUTHOR]

Underwater spectral imaging combined with neural networks provides a practical means for intelligent detection and maintenance of offshore engineering. As an underwater detection technology, the data obtained from underwater hyperspectral imaging can be used for both qualitative analysis and quantitative detection of underwater targets, which is an efficient means of underwater target detection in offshore engineering. In this paper, An underwater hyperspectral imaging system based on PGP spectral structure line scanning imaging has been designed and developed for shallow water detection. It has a spectral range of 400–1000 nm, over 200 channels, a spectral resolution of 2.5 nm, a field of view angle of 35.2°, and a focal length of 25 mm. According to the design results, the system was assembled, calibrated and tested. Aiming at solving the shortcomings of the current underwater hyperspectral system, such as the inability of real-time observation of the scanned image, large data volume and complicated processing, the upper computer software was designed and developed to realize the functions of controlling and adjusting the parameter, real-time displaying the scanned image, reconstructing the image and outputting the data in ENVI standard format to the underwater hyperspectral system. The prototype of the system was used to conduct coral detection experiments in Weizhou Island, and the neural network model was used to classify the experimental data, and the overall classification accuracy was above 98%. The experimental results show that the whole system has good imaging quality and good underwater detection capability, and the developed host computer software greatly reduces the workload of raw data processing, improves the experimental detection efficiency, and provides a feasible technical solution for underwater intelligent detection and analysis. • A line-scanning underwater hyperspectral imaging system has been developed based on the PGP spectroscopic structure. • Development of software for the upper computer, which significantly improves the efficiency of underwater detection. • Combining neural network models to achieve accurate detection and classification of underwater targets. [ABSTRACT FROM AUTHOR]

• The academic publication focuses on applying a range of machine learning (ML) models to forecast CO2 emissions from the building sector until 2050. • The publication compares various ML approaches including linear Regression, ARIMA, and neural Networks, using both univariate and multivariate modelling. • A multi-step methodology is being followed for generating highly accurate forecasts. • The study spans multiple regions, globally, and evaluates the predictive performance using metrics such as the mean absolute percentage error (MAPE). • This is the first application of such a comparison of ML models on datasets related to the building sector's CO2 emissions. The building sector has historically accounted for around 50% of the energy-related carbon dioxide (CO2) emissions on a global scale. As a result, it attracts significant attention as part of the worldwide effort to decarbonize the energy system. This paper presents and compares a variety of Machine Learning (ML)-based approaches for long-term predictions of CO2 emissions from buildings until the year 2050. These approaches include Linear Regression, ARIMA (Autoregressive Integrated Moving Average), Shallow Neural Networks, and Deep Neural Networks, all conducted using both univariate and multivariate modelling and with different approaches for the Features Extraction process; namely, the lagged values approach and the polynomial transformation. The analysis is conducted for different regions of the world including Brazil, India, China, South Africa, the United States, Great Britain, the World-average and the European Union. A variety of tests are conducted to evaluate and compare the predictive performance of the different ML approaches. [ABSTRACT FROM AUTHOR]

Today the tasks of complex artificial and natural objects control have come to the fore in the majority of subject domains. The efficiency and effectiveness of solving these tasks directly depends of the efficiency and effectiveness of data fusion (DF). Data fusion methods are designed to integrate data from multiple sources and transform it in order to produce more consistent, accurate, and useful information than that provided by any individual data source. Although DF has been extensively studied for a considerable period of time it is still hardly applicable in practice in the processes of the control of the real world objects with complex structure and behavior as the data produced by the objects is, in the majority of cases, heterogeneous, multimodal, and imperfect, has huge volume. To ensure proper response to the changes in the state and behavior of the controlled objects that can be caused by both internal and external influencing factors the data should be processed with high accuracy and with minimum delays. Despite the importance of the tasks of complex objects control till now there are no researches that clarify to what extent the DF problem has been solved from the perspective of its application in the processes of objects control based on the data received from the objects. In the survey we define the requirements to DF in the interests of the control of complex artificial and natural objects, consider the structure of the multilevel process of intelligent object control, identify the neural networks that can be used in the control process for data fusion. Despite the wide capabilities of the existing NN we reveal that they still do not meet all the requirements to DF for complex objects control. Based on the analysis of NN architectures, we define requirements for advanced NN architectures and discuss future research directions. To facilitate our literature analyses, we also perform conceptual exploration of collected papers with lattices of closed itemsets and implications from Formal Concept Analysis and Data Mining used for knowledge processing in similar large-scale studies. [Display omitted] • Data fusion with adaptive neural network models for intelligent control of heterogeneous objects. • Continuous neural network learning and operation. • Control of complex nonlinear artificial and natural objects. • Conceptual exploration and data mining of recent literature. • Ethical and practical implications. [ABSTRACT FROM AUTHOR]

Deep Neural Networks (DNNs) are widely used in many application domains. However, they require a vast amount of computations and memory accesses to deliver outstanding accuracy. In this paper, we propose a scheme to predict whether the output of each ReLu activated neuron will be a zero or a positive number in order to skip the computation of those neurons that will likely output a zero. Our predictor, named Mixture-of-Rookies , combines two inexpensive components. The first one exploits the high linear correlation between binarized (1-bit) and full-precision (8-bit) dot products, whereas the second component clusters together neurons that tend to output zero at the same time. We propose a novel clustering scheme based on analysis of angles, as the sign of the dot product of two vectors depends on the cosine of the angle between them. We implement our hybrid zero output predictor on top of a state-of-the-art DNN accelerator. Experimental results show that our scheme introduces a small area overhead of 5.3% while achieving a speedup of 1.2x and reducing energy consumption by 16.5% on average for a set of diverse DNNs. [ABSTRACT FROM AUTHOR]

Despite the recent success of neural network models in mimicking animal performance on various tasks, critics worry that these models fail to illuminate brain function. We take it that a central approach to explanation in systems neuroscience is that of mechanistic modeling, where understanding the system requires us to characterize its parts, organization, and activities, and how those give rise to behaviors of interest. However, it remains controversial what it takes for a model to be mechanistic, and whether computational models such as neural networks qualify as explanatory on this approach. We argue that certain kinds of neural network models are actually good examples of mechanistic models, when an appropriate notion of mechanistic mapping is deployed. Building on existing work on model-to-mechanism mapping (3M), we describe criteria delineating such a notion, which we call 3M++. These criteria require us, first, to identify an abstract level of description that is still detailed enough to be "runnable", and then, to construct model-to-brain mappings using the same principles as those employed for brain-to-brain mapping across individuals. Perhaps surprisingly, the abstractions required are just those already in use in experimental neuroscience and deployed in the construction of more familiar computational models — just as the principles of inter-brain mappings are very much in the spirit of those already employed in the collection and analysis of data across animals. In a companion paper, we address the relationship between optimization and intelligibility, in the context of functional evolutionary explanations. Taken together, mechanistic interpretations of computational models and the dependencies between form and function illuminated by optimization processes can help us to understand why brain systems are built they way they are. [ABSTRACT FROM AUTHOR]

Active learning is focused on minimizing the effort required to obtain labeled data by iteratively choosing fresh data samples for training a machine learning model. One of the primary challenges in active learning involves the selection of the most informative instances for labeling by an annotation oracle at each iteration. A viable approach is to develop an active learning strategy that aligns with the performance of a meta-learning model. This strategy evaluates the quality of previously selected instances and subsequently trains a machine learning model to predict the quality of instances to be labeled in the current iteration. This paper introduces a novel approach to learning for active learning, wherein instances are chosen for labeling based on their potential to induce the most substantial change in the current classifier. We explore various strategies for assessing the significance of an instance, taking into account variations in the learning gradient of the classification model. Our approach can be applied to any classifier that can be trained using gradient descent optimization. Here, we present a formulation that leverages a deep neural network model, which has not been extensively explored in existing learning-to-active-learn methodologies. Through experimental validation, our approach demonstrates promising results, especially in scenarios where there are limited initially labeled instances, particularly when the number of labeled instances per class is extremely limited. • A new meta-active learning strategy for deep neural network models. • The instance selection problem is modeled as a regression problem. • Four strategies of instance importance scoring by considering gradient variations. • Experiments have shown good results when few initially labeled objects are available. • A dimensionality reduction step for the instances improves the performances. [ABSTRACT FROM AUTHOR]

This paper is concerned with the set-membership filtering issue for a class of artificial neural networks subject to mixed attacks. The encoding–decoding communication mechanism is adopted in the processing of data sharing between neurons. During the information exchanges among neurons, both injection and DoS attacks are concurrently considered to reflect the practical operating conditions of the investigated neural networks. The purpose of the addressed problem is to present an algorithm to estimate the neurons' states in the presence of mixed attacks, while guaranteeing the estimation errors at each neuron are confined within certain prescribed ellipsoidal region. Sufficient conditions are derived, in terms of convex optimization approach, to ensure the existence of desired filter, and the explicit filtering parameters are obtained via solving the provided set of matrix inequalities. Finally, a numerical simulation example is proposed to show the validity of the obtained theoretical results. [ABSTRACT FROM AUTHOR]

In this paper, an artificial neural network (ANN) based system-level model predictive control framework is established for a variable air volume (VAV) system to improve its robustness and energy efficiency. The VAV system consists of three processes: the zone temperature process, the damper process and the supply air volume process of the air handling unit connecting the ductwork for all of the boxes. An ANN-based predictive controller is designed for the zone temperature process, and a PI controller is used for the damper process; they operate in a cascaded manner. Another ANN predictive controller tracks the total supply air volume subject to the cooling load constraints from the lower level VAV boxes and minimizes the fan energy consumption. The ANN controllers are optimized online by the Lagrange variational approach (LVA) to minimize the system cost function and implement a feedback receding horizon optimal control. With the LVA based gradient descent training algorithm, the weight convergence of the ANN predictive controllers and stability of the control system is obtained with small computational and storage needs. The modeling methodology, based on ANNs for the above processes, is developed and calibrated with actual data measured from a site. All of the data processing and control strategy implementations are based on extensive measurements collected from a laboratory building located in the campus of Beijing University of Civil Engineering and Architecture , Beijing , China. The experimental results show that an additional 6.12% energy savings is achieved. This is accomplished by introducing the fan control signal into the system cost function as the performance index of energy consumption with a 0.2 penalty coefficient. In the presence of un-modeled disturbances and modeling errors, the framework also reduces the temperature oscillation at least by approximately 60% compared with PI control. • A neural network-based predictive control strategy for VAV systems is designed. • Cascade predictive control is implemented for zone temperature and air flow rate. • The duct air volume process achieves total airflow predictive control with savings in fan energy consumption. • The predictive controller using neural network implementation, easy to achieve nonlinear system optimization. • The control strategy is robust to un-modeled disturbances and modeling errors. [ABSTRACT FROM AUTHOR]

Recent European Community directives introduce Renewable Energy Communities (REC) and Jointly Acting Renewable Self-Consumers (JARSC). Both entities are constituted by communities of residential and/or non-residential prosumers, located in proximity of renewable generators and Electrical Storage Systems (ESS) owned and managed by the REC/JARSCs. These aggregations of prosumers are aimed at providing environmental and economic benefits by maximizing their global self-consumption. In this frame, it is relevant to introduce a control strategy which considers the whole system represented by the REC/JARSCs and performs optimal management of energy production, storage and consumption. The present paper proposes a Model Predictive Control (MPC) based control design, targeted at the minimization of electricity cost and equivalent CO2 emissions, considering the whole ensemble of loads included in the REC/JARSCs over a 24-h prediction horizon. To exploit the MPC ability of including forecasts in the optimization problem, predictors including Artificial Neural Networks (ANN) are developed for solar irradiance, air temperature, electricity price and carbon intensity. The proposed control performance is evaluated considering a case study located in Milan, Italy, and its advantages with respect to traditional control algorithms are highlighted by comprehensive numerical simulations. Lastly, an economic evaluation of the considered system is presented. • RECs and JARSCs are new forms of self-consumption introduced by EU directives. • RECs/JARSCs operation approached as an optimization problem by an MPC Controller. • ANN-based predictors for PV production, electricity price and carbon intensity. • Evaluation of economic revenues and CO2 emission reduction in different Scenarios. [ABSTRACT FROM AUTHOR]

Neural networks have shown tremendous growth in recent years to solve numerous problems. Various types of neural networks have been introduced to deal with different types of problems. However, the main goal of any neural network is to transform the non-linearly separable input data into more linearly separable abstract features using a hierarchy of layers. These layers are combinations of linear and nonlinear functions. The most popular and common non-linearity layers are activation functions (AFs), such as Logistic Sigmoid, Tanh, ReLU, ELU, Swish and Mish. In this paper, a comprehensive overview and survey is presented for AFs in neural networks for deep learning. Different classes of AFs such as Logistic Sigmoid and Tanh based, ReLU based, ELU based, and Learning based are covered. Several characteristics of AFs such as output range, monotonicity, and smoothness are also pointed out. A performance comparison is also performed among 18 state-of-the-art AFs with different networks on different types of data. The insights of AFs are presented to benefit the researchers for doing further research and practitioners to select among different choices. The code used for experimental comparison is released at: https://github.com/shivram1987/ActivationFunctions. [ABSTRACT FROM AUTHOR]

Numerical simulations of combustion processes in rocket engines requires a long run time of supercomputer systems even for a very short physical time. Therefore, creating digital twins of rocket engines needs enormous processor time, and is not very effective. This computational time surpasses the actual physical time of the process in many orders of magnitude. To speed up numerical simulations the paper presents a solution of the chemical kinetics problem using artificial neural network approach. Using the architecture of a multilayer neural network with bypass connections, namely residual network, it is possible to obtain a fast and reliable solution to the problem. The neural network is trained to predict the state of the system only one time step ahead. Using it in a recursive mode, it is possible to forecast for thousands of steps without loss of accuracy. • Simulation of a chemical kinetic problem by the neural network method was carried out. • The network was trained on a wide range of data. • A shortened neural network with preservation of accuracy was obtained. • The considered network used a small amount of RAM (Random Access Memory). [ABSTRACT FROM AUTHOR]

In this paper we present a novel methodology, referenced as ORFEO (Ordinal classifier and Regressor Fusion for Estimating an Ordinal categorical target), to enhance the performance in ordinal classification problems for which the latent variable is observable. ORFEO is an artificial neural network model incorporating two outputs, one for ordinal classification, using the cumulative link model, and one for regression, using a linear model. Both outputs are simultaneously optimised considering a loss function that linearly combines both classification and regression losses. The main motivation behind developing the proposed approach is to enhance the performance of a standard ordinal classifier. This improvement is facilitated by considering the regression output, which allows the model to differentiate between patterns within the same category. The ORFEO model is applied to two problems in the field of marine and ocean engineering: short-term prediction of both significant wave height and flux of energy. Both problems are addressed considering four different coastal zones of the United States of America, using 13 datasets formed by buoys measurements and reanalysis data. A comprehensive comparison against 20 methodologies, including regression and nominal/ordinal classification approaches is performed, by using diverse nominal and ordinal performance metrics. Ranks achieved indicate that ORFEO outperforms all the compared methodologies in terms of all the performance measures, demonstrating the efficacy and robustness of the proposal. Finally, a statistical analysis is conducted, concluding that there are statistically significant differences across ordinal and nominal performance metrics in favour of the proposed ORFEO model. [ABSTRACT FROM AUTHOR]

In the realm of aircraft design, there is a prevalent need for surrogate modeling techniques capable of efficiently and accurately modeling objects with high evaluation costs and uneven sensitivity distributions, such as those encountered in the nonlinear buckling of thin-walled structures and aerodynamics at transonic speeds. However, the non-stationary Gaussian Process Regression (GPR) model requires extensive hyperparameters or judicious assumptions, limiting its applications in such tasks, and the commonly used K-Fold methodology suffers from inherent instability. Addressing these issues, this paper proposes an active learning method for surrogate modeling in systems featuring uneven sensitivity distribution, integrating non-stationary-noise GPR and K-Fold Artificial Neural Network (ANN) methodology, namely NSN-GPR-KF. Instead of relying on a non-stationary kernel function, the algorithm regards noise as a non-stationary factor within a stationary GPR framework, with noise levels estimated via the K-Fold ANN methodology, allowing GPR to efficiently adapt to non-stationary systems while avoiding the aforementioned requirements for non-stationary GPR. Case validations, performed on two test functions with pronounced uneven spatial sensitivity, demonstrate that NSN-GPR-KF outperforms commonly used stationary GPR and pure K-Fold methodology. It synthesizes the exploration capabilities of GPR with the exploitation proficiency of the K-Fold methodology, achieving comparable global accuracy with reduced sample sizes—up to 26 % and 22 % less than the other two algorithms, respectively. The algorithm was successfully applied to predict the buckling load of thin-walled pipe beams, demonstrating accelerated convergence and approximately 25 % greater accuracy compared to pure K-Fold model. These results suggest that NSN-GPR-KF can serve as an efficient and precise modeling tool for engineering tasks with high evaluation costs and complex sensitivity distributions. [ABSTRACT FROM AUTHOR]

• Low-cost measurement platform for particulate matter monitoring developed. • Artificial intelligence field calibration procedure of low-cost sensor proposed. • Combination of affine scaling and optimized neural network surrogate employed. • Correlation coefficient improved to 0.86 for particulate matter of size 1 μm. • Root-mean-squared error of 3 μm/m3 (1 μm particles) and 4.9 μm/m3 (10 μm particles) Due to detrimental effects of atmospheric particulate matter (PM), its accurate monitoring is of paramount importance, especially in densely populated urban areas. However, precise measurement of PM levels requires expensive and sophisticated equipment. Although low-cost alternatives are gaining popularity, their reliability is questionable, attributed to sensitivity to environmental conditions, inherent instability, and manufacturing imperfections. The objectives of this paper include (i) introduction of an innovative approach to field calibration for low-cost PM sensors using artificial intelligence methods, (ii) implementation of the calibration procedure involving optimized artificial neural network (ANN) and combined multiplicative and additive correction of the low-cost sensor readings, (iii) demonstrating the efficacy of the presented technique using a custom-designed portable PM monitoring platform and reference data acquired from public measurement stations. The results obtained through comprehensive experiments conducted using the aforementioned low-cost sensor and reference data demonstrate remarkable accuracy for the calibrated sensor, with correlation coefficients of 0.86 for PM 1 and PM 2.5 , and 0.76 PM 10 (particles categorized as having diameter equal to or less than 1 μm, 2.5 μm, and 10 μm, respectively), along with low RMSE values of only 3.1, 4.1, and 4.9 µg/m3. [ABSTRACT FROM AUTHOR]

Today, a significant content of the system on chips (SOCs) is dedicated to static random access memory (SRAM) cells. Due to the stress during SRAM operation, MOSFET aging is becoming an increasingly serious problem. The manufacturing variability of the MOSFETs causes further differences in the aging of various devices. This paper presents a simulation methodology using a deep neural network (DNN) to predict the performance of SRAM cells, taking into account the impacts of fabrication variations, temperature, and aging. A dataset of four input features and six output criteria is utilized, which are derived from power consumption, power supply transient current, signal rise time, read static noise margin (RSNM), and butterfly and N-curve. The accuracy of the long short-term memory (LSTM) model in predicting the performance of SRAM cells outperformed the other models such as multilayer perceptron (MLP), one-dimensional convolutional neural network (1D-CNN), gated recurrent unit (GRU), and recurrent neural network (RNN) by 1.85 % to 1.11 %. For instance, by processing the supply current data, the LSTM model achieved 97.41 %, 97.96 %, and 97.41 % accuracy in predicting the RMS power consumption, the signal rise time, and the average power supply current of the SRAM cell, respectively. • DNN successfully predicts the aging and process variation impact on SRAM performance. • Supply current illustrates superior results in performance prediction. • LSTM DNN model outperforms MLP, RNN, 1D-CNN, and GRU, exhibiting sufficient accuracy. [ABSTRACT FROM AUTHOR]

• Lotka-Volterra model describes seaport competition for container throughputs. • Nonlinear phenomena are investigated by bifurcation and chaotic behaviors of port operations management. • Adaptive fractional-order control synthesis deals with synchronization for port operation problems. • Proposed approach is successfully validated with case study of Korean ports. • Port authorities cope with market volatility to improve port performance and reliability. This paper is to explore the dynamical analysis and active control synthesis for improving seaport operations through system optimization. First, the multi-dimensional system of the fractional Lotka-Volterra model is utilized to investigate its underlying dynamics of port competition and cooperation for container throughput. The nonlinear dynamical behaviors are intensely examined by using eigenvalue evaluation, bifurcation analysis and time series analysis to illustrate the stability of the competition and collaboration model. With the system theory based on fractional order calculus, various dynamic properties of ports system have been discovered. Based on the dynamical analysis, adaptive fractional-order sliding mode control with artificial neural network algorithm has been realized for port growth rate improvement against disruptions. Furthermore, the proposed approach is successfully validated by the case study of major Korean seaports for improving port operations through neural network prediction. The decision making policies implemented by control algorithms are guaranteed to improve the port growth rate against disruptions. Specifically, the novel management strategy is to provide managerial insights and innovative solutions for seaport authorities who can make more efficient strategic planning for handling risk management in maritime logistics. [ABSTRACT FROM AUTHOR]

This paper focuses on an adaptive learning and control problem for a class of discrete-time nonlinear uncertain systems operating under multiple environments. A novel intelligent learning control framework is proposed by using a combination of offline and online learning methods. Specifically, in the offline learning mode, a deterministic learning (DL) based adaptive dynamics learning approach is first proposed to achieve locally-accurate identification of associated nonlinear uncertain system dynamics under each anticipated individual environment, and the learned knowledge is obtained and stored in a set of constant radial basis function neural network models. Then, with the learned knowledge, an online adaptive learning control scheme is further developed, which consists of: (i) an online adaptive learning control mechanism composed of multiple experience-based controllers and a DL-based adaptive learning controller, aiming to provide desired control performance for the plant operating under each individual environment; and (ii) a learning-based recognition mechanism composed of multiple recognition estimators and a DL-based identifier, aiming to recognize the active environment and schedule appropriate control strategies in real time. To guarantee the system stability during environment transition, a robust quasi-sliding-mode controller is further developed and embedded in the overall controller architecture. With this new intelligent adaptive learning control framework, the overall system is capable of adapting not only to any anticipated (pre-defined) environment by re-utilizing the knowledge obtained from both offline and online learning, but also to unanticipated (new) environments by actively acquiring new knowledge online. Simulation studies are conducted to verify the effectiveness and advantages of this new framework. [ABSTRACT FROM AUTHOR]

Accurate forecasting of regional solar photovoltaic power (SPVP) generation is essential for efficient energy management and planning. Existing approaches have shown the effectiveness of decomposing the time series to model the stochastic variability in SPVP data. However, these approaches have limitations in extracting and exploiting both spatial and temporal information from complex and high-dimensional data from multiple sources with intricate relationships, which can impact the accuracy of predictions. In this paper, we propose a novel approach called multilevel data fusion and neural basis expansion analysis (MF-NBEA) for forecasting aggregated regional-level SPVP generation. MF-NBEA integrates exogenous data at multiple levels, uses supervised and unsupervised encoders to provide compact data representation, and enhances model learning from complex data by incorporating spatial information. It also includes a sequence analyser module based on a neural network decomposition mechanism to learn the variability in data and incorporates a residuals learner module to improve overall predictions. We evaluate MF-NBEA using two real-world datasets and find that it outperforms state-of-the-art deep learning methods in terms of forecast accuracy. Furthermore, MF-NBEA facilitates information fusion and knowledge extraction to provide interpretable predictions regarding trend, seasonality, and residual components. The insights gained from our approach inform decision-making for energy management and planning, and can lead to more efficient and sustainable resource utilisation. • MF-NBEA for multivariate solar power time series forecasting. • A 3D autoencoder encodes spatial and temporal information from heterogeneous sources. • Interpretable forecasting in terms of trend, seasonality, and residuals. [ABSTRACT FROM AUTHOR]

The burner combustion tuning is a complex problem that has been studied through flame monitoring and characterization. It has been observed that the flame electromagnetic spectrum and flickering contain specific flame information in combustion processes. This information is helpful for combustion stoichiometry tuning on burners. This paper described a method for selecting the best flame feature subset that can be computed from the scanner signal, in order to get the flame index and induce combustion stoichiometry on burners under specific combustion conditions. We propose a method for selecting a reduced subset with only the useful flame features for flame index classification. To extract the most relevant flame features we use a feature subset selection (FSS) algorithm and to determine the combustion state in burners, five flame indices were defined that represent the most common flame states in oil fuel-fired boilers. FSS includes complete, sequential, and random searches in order to eliminate redundant and noisy flame features to decrease the flame feature set dimension. A probabilistic neural network (PNN) algorithm was implemented for flame feature clustering. Signals from the actual flame scanner system and relevant variables from the boiler data acquisition system were used by the algorithms to calculate the burner flame index. A set of parametric tests was done in a heavy oil-fired boiler under well-known flame and index conditions to train and test the flame classifier. The results showed that only the four more relevant features are enough to classify flames with a good performance (92.3% accuracy), which is useful for burner combustion monitoring and optimization. • A new approach to defining/selecting critical features in the combustion process. • A new approach to classifying flames in industrial heavy oil-fired boilers. • The proposed algorithms are tested & evaluated in a real industrial process. • The overall methodology ensures that the combustion process guarantees safety. [ABSTRACT FROM AUTHOR]

As the "respiratory tract" of the air breathing engine, the hypersonic inlet plays the role of compressing, decelerating, and pressurizing flow. However, research on the traditional inlet design is mainly based on gas dynamics theory and experience, which not only leads to a long design period but also fails to ensure the performance of the inlet at non-design points under broad working conditions. In recent years, with the further development of flow physics, a variety of advanced inlet design methods have been proposed, and a large number of valuable data have been accumulated by applying high-fidelity computational fluid dynamics numerical simulation and ground wind tunnel tests. A new generation of machine learning technologies, typically represented by deep learning, is developing vigorously. By building a deep neural network structure and using data-driven methods to carry out model training, it can realize typical feature extraction automatically, efficiently, and accurately, which is helpful to deeply explore the hidden flow mechanism between data and establish a fast prediction model of inlet performance. The optimal inlet design scheme can be obtained by applying the performance intelligent prediction model and the multi-objective intelligent evolution algorithm. This paper provides a detailed overview of the latest progress in inlet design by applying traditional ideas, expounds the basic principles and typical applications of some representative and prospective machine learning methods, and especially reports the current research status of the combination of machine learning and inlet. Finally, the future development trends and potential applications of several research directions are summarized. [ABSTRACT FROM AUTHOR]

Artificial neural networks, inspired by the biological networks of the human brain, have become game-changing computing models in modern computer science. Inspired by their wide scope of applications, synthetic biology strives to create their biological counterparts, which we denote synthetic biological neural networks (SYNBIONNs). Their use in the fields of medicine, biosensors, biotechnology, and many more shows great potential and presents exciting possibilities. So far, many different synthetic biological networks have been successfully constructed, however, SYNBIONN implementations have been sparse. The latter are mostly based on neural networks pretrained in silico and being heavily dependent on extensive human input. In this paper, we review current implementations and models of SYNBIONNs. We briefly present the biological platforms that show potential for designing and constructing perceptrons and/or multilayer SYNBIONNs. We explore their future possibilities along with the challenges that must be overcome to successfully implement a scalable in vivo biological neural network capable of online learning. [ABSTRACT FROM AUTHOR]

Activation functions play an important role in deep learning models by introducing non-linearity to the output of a neuron, enabling the network to learn complex patterns and non-linear relationships in data and make predictions on more complex tasks. Deep learning models' most commonly used activation functions are Purelin, Sigmoid, Tansig, Rectified Linear Unit (ReLU), and Exponential Linear Unit (ELU), which exhibit limitations such as non-differentiability, vanishing gradients, and neuron inactivity with negative values. These functions are typically defined over a finite range, and their outputs are integers or real numbers. Using fractional calculus in designing activation functions for neural networks has shown promise in improving the performance of deep learning models in specific applications. These activation functions can capture more complex non-linearities than traditional integer-order activation functions, improving performance on tasks such as image classification and time series prediction. This paper focuses on deriving and testing linear and non-linear fractional-order forms of activation functions and their variants. The linear activation function includes Purelin. In contrast, the non-linear activation functions are Binary Step, Sigmoid, Tansig, ReLU, ELU, Gaussian Error Linear Unit (GELU), Hexpo, and their variants. Besides, the standard formula has been implemented and used in developing the fractional-order linear activation function. Furthermore, various expansion series, such as Euler and Maclaurin, have been used to design non-linear fractional-order activation functions and their variants. The single- and multi-layer fractional-order neural network models have been developed using the designed fractional-order activation functions. The simulation study uses developed fractional-order neural network models for predicting the Texas wind turbine systems' generated power. The performance of single and multi-layer fractional-order neural network models has been evaluated by changing the activation functions in the hidden layer while keeping the Purelin function constant at the output layer. Experiments on neural network models demonstrate that the designed fractional-order activation functions outperform traditional functions like Sigmoid, Tansig, ReLU, ELU, and their variants, effectively addressing limitations. • Activation functions are derived into fractional functions using fractional calculus. • The fractional neural network models have been made by adopting derived functions. • A Case study has been done on the Texas wind turbine with the models to predict the power. • The models' performance has been evaluated using the fractional functions. [ABSTRACT FROM AUTHOR]

This paper proposes a computationally efficient and effective data-driven modeling framework for dynamical systems. The proposed modeling framework employs a collection of shallow neural networks known as Extreme Learning Machines (ELMs) to model local system behaviors along with data-driven inferred transitions among local models to establish a neural hybrid automaton model. First, the sampled system inputs are mapped to the corresponding feature spaces to obtain data-driven partitions, which subsequently define the transitions and invariants of the neural hybrid automaton model through a novel data-driven mode clustering process. Then, a collection of ELMs are trained to approximate the local dynamics. The learning processes integrate a segmented data merging procedure for location identification and a local dynamics modeling process. The proposed neural hybrid automaton models can capture behaviors of complex dynamical systems with high modeling precision but significantly lower computational complexities in computationally expensive tasks such as training and verification, which are traditionally considered to be computationally expensive tasks for neural network models. A computationally efficient set-valued reachability analysis method which is commonly used in safety verification is then developed based on interval analysis and a novel Split and Combine process. Finally, applications to modeling the limit cycle and human handwritten motions are presented to show the effectiveness and efficiency of our approach. [ABSTRACT FROM AUTHOR]

The emergence of a variety of Machine Learning (ML) approaches for travel mode choice prediction poses an interesting question to transport modellers: which models should be used for which applications? The answer to this question goes beyond simple predictive performance, and is instead a balance of many factors, including behavioural interpretability and explainability, computational complexity, and data efficiency. There is a growing body of research which attempts to compare the predictive performance of different ML classifiers with classical Random Utility Models (RUMs). However, existing studies typically analyse only the disaggregate predictive performance, ignoring other aspects affecting model choice. Furthermore, many existing studies are affected by technical limitations, such as the use of inappropriate validation schemes, incorrect sampling for hierarchical data, a lack of external validation, and the exclusive use of discrete metrics. In this paper, we address these limitations by conducting a systematic comparison of different modelling approaches, across multiple modelling problems, in terms of the key factors likely to affect model choice (out-of-sample predictive performance, accuracy of predicted market shares, extraction of behavioural indicators, feature importance analysis, and computational efficiency). The modelling problems combine several real world datasets with synthetic datasets, where the data generation function is known. The results indicate that the models with the highest disaggregate predictive performance (namely Extreme Gradient Boosting (XGBoost) and Random Forests (RF)) provide poorer estimates of behavioural indicators and aggregate mode shares, and are more expensive to estimate, than other models, including Deep Neural Networks (DNNs) and Multinomial Logit (MNL). It is further observed that the MNL model performs robustly in a variety of situations, though ML techniques can improve the estimates of behavioural indices such as Willingness To Pay (WTP). [Display omitted] • Systematic evaluation of ML and discrete choice models for travel mode choice. • Addressing common methodological failures found in the literature. • Assessing the quality of model outputs extracted from ML methods. • Evaluating the models on a wide variety of real-world and synthetic datasets. [ABSTRACT FROM AUTHOR]

Though Deep Neural Networks (DNNs) have been widely deployed and achieved great success in many domains, they have severe safety and reliability concerns. To provide testing evidence for DNNs' reliable behaviors, various coverage testing techniques inspired by traditional software testing have been proposed. However, the coverage criteria in these techniques are either not fine enough to capture subtle behaviors of DNNs, or too time-consuming to be applied on large-scale DNNs. In this paper, we propose a coverage testing framework named HashC, which makes mainstream coverage criteria (e.g., NC and KMNC) much finer. Meanwhile, HashC reduces the time complexity of combinatorial coverage testing from polynomial time to linear time. We also develop the corresponding test sample selection method. Our experiments show that, (1) the existing mainstream coverage criteria are becoming finer after being equipped with HashC, (2) HashC greatly accelerates combinatorial coverage testing and can handle the testing of large-scale DNNs. [ABSTRACT FROM AUTHOR]

Various microreactor designs under development aim at fulfilling electricity and heat production requirements affordably and reliably in a variety of applications, like power generation for remote communities and military bases. The fission battery deployment strategy for microreactors provides battery-like simplicity for its installation and operation. A successful fission battery must be cost competitive, have unit standardization, enable flexible installation, possess high reliability, and feature unattended operations. Due to the design and operational features of microreactors, successful unattended operations of the system must be enabled through an on-line monitoring (OLM) system that can effectively and reliably detect and diagnose any fault in the reactor. This research introduces the concept and approach of an intelligent Fault Detection and Diagnosis Monitoring System (FDDMS) that can provide power transient dependent fault detection and diagnosis to fit within a future autonomous control framework. Specifically, this paper develops the first FDDMS module, the Power Transient Module, to reliably identify reactor operations as steady state, ramping up in power, or ramping down in power using data from a broad-scope simulator. Knowledge of the current power transient in the reactor is paramount for effective fault detection and diagnosis in any operational regime. This paper explores various data-driven methods to accurately evaluate the power transients: 1) Principal Component Analysis and Support Vector Machines, 2) Deep Neural Networks, and 3) Convolutional Neural Networks. To prepare the sensor data collected from the simulator for training the data-driven models, several preprocessing techniques were tested to optimize for evaluation accuracy. The final model selected for the Power Transient Module produced perfect evaluation results for several power transients, including in previously unseen regimes. Foundation for further module development of the FDDMS was established. Future works will see the implementation of the remaining FDDMS architecture. • Fault Detection and Diagnosis Monitoring System introduced to provide basis for reactor operations with minimal to no personnel on-site. • Several data-driven models tested and compared to accurately evaluate different power transients present in reactor operations. • Several data transformation methods explored to create 2D representations from sensor data to illustrate multi-modal temporal relationships. [ABSTRACT FROM AUTHOR]

Domain adaptation (DA) methods facilitate cross-domain learning by minimizing the marginal or conditional distribution shift between domains. However, the conditional distribution shift is not well addressed by existing DA techniques for the cross-domain regression learning task. In this paper, we propose Multi-Domain Adaptation for Regression under Conditional shift (DARC) method. DARC constructs a shared feature space such that linear regression on top of that space generalizes to all domains. In other words, DARC aligns different domains according to the task-related information encoded in the values of the dependent variable. It is achieved using a novel Pairwise Similarity Preserver (PSP) loss function. PSP incentivizes the differences between the outcomes of any two samples, regardless of their domain(s), to match the distance between these samples in the constructed space. We perform experiments in both two-domain and multi-domain settings. The two-domain setting is helpful, especially when one domain contains few available labeled samples and can benefit from adaptation to a domain with many labeled samples. The multi-domain setting allows several domains, each with limited data, to be adapted collectively; thus, multiple domains compensate for each other's lack of data. The results from all the experiments conducted both on synthetic and real-world datasets confirm the effectiveness of DARC. • We propose a new pairwise similarity preserver loss function for regression tasks. • We propose a new domain adaptation method for regression tasks. • Our proposed method adapts domains with any arbitrary shift, concept, or covariate. • Our proposed method is capable of adapting multiple domains. • The results prove our method's efficiency in solving real-world problems. [ABSTRACT FROM AUTHOR]

• Development of a prediction model for airborne chloride depositions with surrounding environmental information. • Bayesian regularization-based artificial neural network considering the high data variance. • Average 35% performance improvement with Bayesian regularization compared to conventional ANN. This paper suggests an artificial neural network model combining Bayesian regularization (BRANN) to estimate concentrations of airborne chlorides, which would be useful in the design of reinforced concrete structures and for estimating environmental effects on long-term structural performance. Meteorological and topographical data were collected, and airborne chlorides were measured at 19 areas all over Korea. Data were classified for the three major coasts, and then prepared for training. To investigate the relationship between each input feature and output and then construct a model for estimating airborne chlorides with only meteorological and topographical information, both the standard artificial NN (ANN) and BRANN models were examined. The 3 or 4-layered BRANN model with 64 nodes at each layer showed the best and most robust performance. This BRANN model successfully predicted airborne chloride content with reasonable values of MSE and R-square although the input data and the airborne chlorides had quite low correlation. The results showed that the BRANN was better able to solve this problem than ANN. It is expected to be broadly applicable for predicting the penetration of chlorides into concrete and even the evaluation of concrete durability. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of generating controllers for automatic bucket filling using data collected during normal operations by professional operators on worksites. We introduce methods to annotate different phases of bucket fillings from long time series data and synthesize five machine learning models. The methods learn a policy that maps input signals to control commands using imitation learning from the training data. We use four different datasets as training data and first compare the model's accuracy on predicting operator commands. We then deploy and evaluate the efficacy and robustness of the policies on a real machine. The experiments shows that the Multilayer Perceptron and the Convolutional Neural Network models had the best overall performance and robustness and could achieve human level performance. Furthermore, models with high prediction accuracy are not necessarily suitable for feedback control, and using data from real worksites increased the controller's robustness. • Automatic labeling of wheel loader bucket fillings from time series data. • Evaluating machine learning models with different datasets collected from worksites. • Human level performance for selected synthesized machine learning controllers. [ABSTRACT FROM AUTHOR]

Useful information for offshore systems can be obtained by monitoring sea state parameters, but the accurate estimation of the on-site wave condition is challenging. This paper deals with motion-based wave estimation using neural networks. The main novelty of this contribution is the evaluation of the performance of the data-driven inference system with motion data from different draft conditions. The study is based on the simulated motions of a Floating Production Storage and Offloading (FPSO) unit with multiple loadings, subjected to typical sea conditions observed offshore the Brazilian coast. The inference models were trained to estimate two independent sets of wave related parameters, corresponding to modal wave statistics, which were post processed to identify single and double-peaked sea states. Results showed that estimation errors were quite similar along the evaluated drafts, with an overall good agreement between estimations and reference values. However, a limitation of the method regarding less frequent sea conditions was observed, even when significant wave response is induced. Also, an analysis of the robustness of the estimation models with respect to a draft variation of ± 0.5 m was carried out considering the global wave statistics, presenting small average deviations compared to the expected values. • Ship motion-based wave inference is performed with neural network models. • Crossed sea states are considered by estimating two sets of sea spectrum parameters. • Models derived from data of different vessel drafts present similar performance. • The inference models are robust for a variation of ± 0.5 m in the draft condition. • Less frequent sea conditions may present challenges to the data-driven estimation. [ABSTRACT FROM AUTHOR]